CodeLlama 70B
MetaCode Models
79.0
Performance
★ 4.2
Rating
310
Reviews
Code AssistantOpen WeightMedium
About
Meta's largest code-specialized Llama model, fine-tuned for code generation, completion, and infilling.
Strengths
Strong open-weight code model supporting Python, JavaScript, C++, Java, and many more languages. Code infilling support for IDE integration. Can run on consumer hardware with quantization. Llama license for commercial use.
Specifications
- Context window
- 100,000
- Parameters
- 70B
Speed & Latency
- 50
- tokens/sec
- 350ms
- time to first token
Available On
HuggingFaceTogether AIFireworks AIOllama
Features
code completionfill in the middlestreamingmulti language
Performance Trend
Benchmark score trends over time for the top 5 benchmarks.
Loading history...
Benchmarks
Scores from various benchmark tests; higher is better.